As defense systems continue to shrink, corresponding thermal management concerns expand exponentially. Designers have learned that overheating can be the downfall of even the most well-designed systems. Suppliers of today’s box-level systems continue to make strides in reducing the size, weight and power (SWaP) of these systems to meet mil-aero deployment demands. However, smaller systems may be more difficult to cool, adding an additional design mandate to keep the system within specified heat parameters, making thermal management that much more important in meeting environmental deployment specifications. Therefore, optimized cooling techniques need to be employed to meet SWaP-C (Size, Weight, Power and Cooling) needs that are now a vital part of the development process.
Designing ruggedized systems for high mission-critical reliability requires testing and validation employing sophisticated thermal modeling tools and computational fluid dynamics (CFD) evaluation techniques to accurately predict airflow, temperature distribution, and heat transfer in components, boards and ultimately the complete system. These tools must take many thermal methodologies into consideration, such as optimal fin geometries based on the internal system layout along with the component conduction path. This article covers the main critical focus areas in box-level computing platforms that need to be solved with thermal optimization.
Thermal Design Approaches
The thermal design approach that has proven most effective over time is to implement all of the required system functionality in a chassis that has been pre-certified for ruggedized operation in contrast to trusting a chassis that is categorized as “designed to meet.” Systems that have been manufactured and validated to meet the various environmental requirements of MIL-STD-810G give developers assurance of their ability to withstand specified extremes of temperature, vibration, shock, salt spray, sand and chemical exposure. This way, the system is certified to maintain a sealed and temperature-controlled environment protecting and ensuring the reliability of the electronics inside.
Another thermal management consideration is to evaluate the possible effects of radiative cooling in passively cooled convection systems that operate at low power. The size, weight and power reductions in military electronics cause radiation to have a significant impact on where components can be placed or where the completed system can be deployed. A frequent remedy for radiationimpacted systems is to look for a sealed system design that uses a natural convection approach that delivers both scalability and excellent power dissipation.
Designing for ruggedization requires testing and validation employing sophisticated thermal modeling tools and CFD evaluation techniques to accurately predict airflow, temperature distribution, and heat transfer in components, boards, and ultimately the complete system. Consider that a typical small form factor aluminum chassis, where the enclosure is 15 to 20oC over ambient, may dissipate up to one third of its power through the effects of radiation. This is a significant proportion of overall power dissipated and can become even more significant at the higher altitudes experienced by UAVs for example.
Evaluating Thermal Optimization Techniques
For the thermal optimization of rugged box-level systems, there are four primary focus areas to explore and ultimately solve:
- Computational fluid dynamic (CFD) driven parametric optimization of the enclosure fin interface with the ambient environment.
- A system level thermal analysis to determine the power dissipation trade-offs and impacts of one internal sub-system versus another.
- Exploration of primary internal thermal conduction paths to the enclosure.
- An evaluation of installation platform thermal contributors to overall system performance.
The design of the cooling fin geometry is a good first step when perfecting the thermal performance of a natural convection cooled product. For example, a baseline design that has proven to work well is to incorporate finning formed into the enclosure housing upper surface. This design takes away as much heat as possible from the circuit board and processor, which is typically placed just underneath the housing surface. The upper surface fins can be supplemented if needed with various sized modular rear and side wall mounted heat sinks.
With multiple fin design parameters to evaluate in addition to the four finned surfaces on the enclosure, the frequently used iterative and general understanding approach typically takes too much valuable development time and may not result in meeting the application’s performance goals. Today, there are new advanced software tools that allow the designer to get more detailed data than what is offered with traditional CFD software tools. For example, there is a design optimization tool that compliments the existing ANSYS Icepak CFD software that uses powerful algorithms to evaluate sensitivities against multiple variables to guide the engineer to the most advantageous fin geometries for a given design.
Through the careful application of these new tools and analyzing a large “Design of Experiments” suite of design scenarios, these sophisticated software tools streamline the task of determining the fin geometry best suited for a specific application environment in the fewest number of iterations. It is still always wise to verify these fin design parameters in the final CFD analysis. There are graphs that show the relative heat dissipated from the processor when various fin parameters are plotted (thinner vs. thicker; fin spacing) and which direction the design should take next.
The other sub-systems inside the enclosure also need to be evaluated. From this evaluation, designers can best determine what design trade-offs, if any, are required. A typical trade-off evaluation can include looking at the power dissipated by an optional XMC expansion card, and the potential rise in operating temperature from the processor on a computer-on-module board in close proximity. This is where CFD tools can be used to test the thermal relationships between the various electronic modules.
As most companies have finite element analysis, using CFD tools to test likely areas that can cause heat problems becomes invaluable information. Key subsystem areas to look at include component maximum operating temperatures, low and high temperature processor thresholds, power and power density of components, and sidewall versus internal to external wall conductive path components. All these are important when selecting an optioned product profile for a specific application, as well as to maintain compliance with customerspecified MIL or RTCA test standards.
Internal Conduction Path Optimization
The solid state conduction paths from high-power components inside the enclosure must also be confirmed to make sure there are efficient thermal paths to the enclosure walls. To accomplish this, thermal simulation tools inside the CAD design software can be leveraged to find an optimal solution. As an example, the width of heat spreader may need to be changed to optimize the gradient thermal path to the top surface. The final solution should provide minimal thermal resistance while at the same time maintain a low mass to be most effective.
Operational Platform Thermal Factors
The last area to evaluate is operational environment where the intended box-level platform and full system will be deployed. Local environmental factors can considerably impact the “as installed” performance.
So how close to the performance envelope is the system and platform if it needs to be deployed in, for instance, thinner air conduction conditions? Thus, the mounting platform material, mounting orientation, vicinity to other electronic equipment, altitude and potential solar loading are all factors that require careful consideration. A system on a UAV may have the benefit of a colder environment but there still is the thinning atmospheric effects at higher altitudes to consider that can affect, for example, cooling fans.
The two relevant examples here show typical operational factors to evaluate for thermal optimization. The first considers the impact of mounting a small box-level form factor platform on an aluminum cold plate. The second considers the radiation exchange of neighboring electronic enclosures of similar power.
These evaluations spotlight the need for a thorough understanding of contributing thermal factors to select the right solutions that address the issues and ensure reliable operation of the system in the “as installed” environment. For the most part, defense system engineers are innovative in finding creative cooling solutions to complex problems and are expert at designing for worst case scenarios.
Thermal Optimization Makes Trusted Systems
What doesn’t seem to change is that expectations of current and next-generation defense systems continue to expand, calling for ever more integrated capabilities from increased computational performance and communication bandwidths that equate to greater power consumption, resulting in higher heat generated. These same systems must also meet SWaP-C and extreme ruggedization requirements. While currently available boxlevel computing systems have advanced to meet these requirements, it cannot be underestimated that careful consideration must be made to maintain mission-critical reliability by effectively handling many thermal issues — from the ambient environment and inherent component/subsystem power dissipation to additional “as installed” thermal factors. A comprehensive understanding of these contributing thermal factors will ensure the reliable operation of a deployed rugged system.
This article was written by Simon Parrett, Conceptual / Structural / Thermal Engineer, Kontron (Poway, CA). For more information, Click Here .